Health Stream Issue 21 March 2001

The US EPA has announced its intention to withdraw the revised Arsenic Rule only weeks after the Rule was published in the Federal Register on 22 January 2001. This decision follows a review of the legislation by the incoming Bush government, which had earlier announced its intention to scrutinise a number of decisions made in the last few weeks of the Clinton administration.

The US Congress had previously extended the deadline for promulgation of the Arsenic Rule to June 2001, in order to provide a full 12 months for review and debate as requested by the American Water Works Association (AWWA) and other stakeholders, however the EPA decided to press ahead with publication prior to the change of government.

The revised Rule lowered the arsenic standard for drinking water from 50 ppb (parts per billion or micrograms per litre) to 10 ppb, and was expected to affect about 3,000 community water systems and 1,100 nontransient, noncommunity water systems with arsenic concentrations exceeding the new limit. Most affected systems are located in the western states, and in parts of the mid-west and New England. About 97 percent are small systems, serving less than 10,000 people each.

The 10 ppb standard was chosen by the EPA after consideration of the potential health benefits, water treatment costs and technical feasibility for levels of 3, 5, 10 and 20 ppb. The formulation of the Rule was surrounded by controversy, with heavy criticism of the EPA's interpretation of the scientific evidence on health risks, and the validity of assumptions and lack of transparency of the economic calculations (1). While the EPA estimates that the maximum compliance cost will be about US$327 per year for households in small communities, the AWWA believes the costs will be 3 to 4 fold higher.

Soon after the Arsenic Rule was promulgated, the cities of Albequerque (New Mexico) and El Paso (Texas), and the states of Nebraska and New Mexico filed notice of their intent to sue the EPA over the legislation. Law suits were also lodged by several organisations including the Western Coalition of Arid States (WESTCAS), the National Miners Association, and the Wood Preservers Institute. The legal actions challenge the Rule on the basis that the EPA has failed to follow the directive of the US Congress to base Safe Drinking Water Act regulations on the best available science.

The president of WESTCAS (representing water and wastewater agencies in seven western states) stated that "by using faulty science, EPA exaggerated the benefits and minimised the costs the new standard will impose on water customers throughout the country and particularly in the west". The law suits lodged by the National Miners Association and the Wood Preservers Institute also charge that the EPA failed to consider the implications of the new arsenic standard for their industries.

The American Water Works Association, also a strong critic of the EPA benefit-cost analysis, decided not to take legal action against the new Rule but called for the US Congress to ensure that adequate federal funds were available to assist small communities to comply with the new standard.

In the 20 March media release announcing its intention to withdraw the Arsenic Rule, the EPA stated that independent reviews would be commissioned of both the science underlying the 10 ppb standard, and the cost estimates for compliance with the new level.

British researchers recently published the results of genotyping over 2,000 isolates of Cryptosporidium parvum from human and animal sources in the UK ⁽¹⁾, which showed differences in the geographic and seasonal distribution of the two major genotypes of this pathogen. The isolates were genotyped by extracting DNA from oocysts and using PCR (polymerase chain reaction) to amplify several genes including two 18s ribosomal RNA fragments, the COWP gene (Cryptosporidium oocyst wall protein), and the TRAP-C1 and TRAP-C2 genes (thrombospondin-related adhesive proteins). The amplified DNA segments were then characterised by restriction enzyme digestion and/or DNA sequencing. This process permitted the identification of isolates of Type 1 (which infects only humans) and Type 2 (which infects both humans and other mammals including cattle and sheep)..

Characterisation of 71 isolates from calves and lambs showed that all belonged to Type 2, as expected. Analysis of 1,300 isolates from sporadic human infections (cases not related to known outbreaks) showed that 833 (64%) contained oocysts of Type 2, 442 (34%) contained Type 1, 13 (1%) had a mixture of Type 1 and Type 2, and 12 (1%) had a different genotype (designated as Type 3). Further characterisation of Type 3 isolates suggested that these were strongly related to C. meleagridis, a species previously isolated in turkeys.

When 706 isolates related to seven drinking water outbreaks were examined, the Type 1 genotype predominated in 4 outbreaks (indicating human faecal contamination of water), while Type 2 was predominant in the remaining three outbreaks (all with evidence of water contamination by sheep faeces during the spring lambing season). The results for 51 human isolates from 5 swimming pool outbreaks showed a mixture of Type 1 and Type 2 in three outbreaks, with one of the remaining outbreaks attributed to Type 1, and the other to Type 2.

Previous work by the same authors on a smaller collection of 978 sporadic cases collected in England between August 1998 and November 1999 showed marked differences in seasonal patterns between the two genotypes ⁽²⁾. Type 1 was rare in the late-winter-spring season (Feb to May) and peaked in late-summer-autumn (Aug to Oct). In contrast, Type 2 was fairly common all year round, with a peak in spring and continuing high levels throughout summer and autumn.

There was also considerable geographic diversity when the distribution of the genotypes in different regional health areas was compared. In some areas, there was a heavy predominance of Type 2 (up to 87%) while others had equal amounts of Type 1 and Type 2. Overall, about 11% (103/978) of cases were associated with recent foreign travel, with about two-thirds of these being of Type 1. The authors suggest that introduction of Type 1 strains by foreign travellers and subsequent secondary spread from person-to-person and via drinking water, recreational water or food may be important factors in maintaining this genotype in the UK population. For Type 2 strains, livestock form a permanent reservoir from which infections may spread to humans by direct contact, or contamination of drinking water, recreational water or food. Infections of this genotype may also be maintained within the human population by person-to-person spread.

While a number of genetic markers exist which permit the distinction of Type 1 and Type 2 strains of C. parvum as described in the studies above, there are no established methods to differentiate isolates within each genotype. Such techniques to identify individual isolates are highly desirable to further understand the epidemiology of this pathogen and allow the importance of different routes of transmission to be better characterised. Researchers at the Centers for Disease Control and Prevention in the US have now identified a promising genetic marker which may fulfil this purpose ⁽³⁾.

The marker is a small segment of double stranded RNA which appears to be carried by all isolates of C. parvum. Two such double stranded RNAs were recently identified in C. parvum. These are not considered to be part of the normal genetic material (C. parvum chromosomes are made of DNA), but may be defective remnants of viruses which are still able to replicate and be inherited in a stable manner, but do not harm the cell.

A 173 nucleotide segment of the smaller double stranded RNA from 23 cattle isolates and 38 human isolates was sequenced and the results compared. All of the isolates had previously been classified as Type 1 or Type 2 using other genetic probes. Eighteen different RNA sequences were identified among the 61 isolates. The 38 Type 1 isolates could be grouped into 10 subgenotypes and the 23 Type 2 isolates could be grouped into 8 subgenotypes on this basis. Examination of groups of isolates from identified outbreaks showed that all isolates from an individual outbreak were of the same subgenotype. Some subgenotypes occurred in several geographically distinct isolates, indicating they had a widespread distribution in the environment.

These preliminary results suggest that this segment of the double stranded RNA fragment may be a useful marker to track individual C. parvum isolates, as it provides a higher level of discrimination than is currently available using other genetic markers. This would greatly facilitate outbreak investigations, and epidemiological studies seeking to clarify the importance of different transmission routes in human disease.

(2) McLauchlin J et al. (2000). Molecular epidemiological analysis of Cryptosporidium spp. in the United Kingdom: results of genotyping Cryptosporidium spp. in 1,705 fecal samples from humans and 105 fecal samples from livestock animals. Journal of Clinical Microbiology 38(11):3984-3990.

Legionella bacteria are naturally occurring inhabitants of moist environments, and are sometimes found in low numbers in drinking water supplies. Over 20 species of Legionella have been identified, with Legionella pneumophila being the most important human pathogen. Infection with Legionella species can result in either of two forms of illness; Pontiac fever, an acute flu-like illness which last about 1 week and has no long term effects, or Legionnaires’ disease, a form of pneumonia which may be fatal particularly in the elderly or those with underlying illness.

Infection occurs by the inhalation of water aerosols containing the bacteria, and can not be acquired from drinking contaminated water or from person-to-person spread. Legionnaires’ disease was first recognised in 1976, when an outbreak occurred at a convention of the American Legion in Philadelphia, however analysis of stored clinical samples from earlier outbreaks of pneumonia with unidentified causes suggest that the disease had been occurring as early as 1957.

Legionella bacteria can survive in water temperatures between 0°C and 63°C, and grow readily at temperatures between 32°C and 35°C. They are also harboured within intracellular vesicles in some amoeba species, and this is believed to contribute to their survival under adverse environmental conditions. Environments favourable to Legionella growth include cooling towers and evaporative condensers, hot water systems below 60°C and spa baths. Cooling towers, because of their potential to spread aerosols over extensive areas, generally tend to be associated with larger outbreaks than other sources. The public health management approach to this problem has been to require regular maintenance, cleaning and disinfection of towers.

The role of specific tests for Legionella in the management of cooling towers, rather than testing for general indicators of bacterial growth (such as heterotrophic plate count) is an area of controversy. A study of Legionella numbers in cooling towers suggested that those which had been linked to disease outbreaks were more likely to contain high numbers of Legionella, than those not linked to outbreaks, and suggested that high colony counts (>1,000 colony forming units /ml) be used as an indicator of high risk to trigger decontamination procedures ⁽¹⁾.

Recently published Australian research has shown that numbers of Legionella bacteria in cooling towers vary markedly over short time intervals, and that in most cases there was no significant relationship between the colony counts obtained in water samples only one week apart ⁽²⁾ . About one third of towers had elevated counts (in the high risk category) at least once during the summer season, suggesting that classification on the basis of a single sample is unlikely to be a valid indication of risks.

The study was undertaken on 31 cooling towers over a 16-week period in the summer season. All towers were maintained in compliance with Australian Standard AS 3666:1989, and water samples were taken twice weekly at the same time of day and processed within 4 hours. Data from 3 towers was excluded because of variations from standard operating procedures or extended shut down periods.

The 28 towers showed wide variations in mean Legionella count from as low as 2 CFU (colony forming units)/ ml to as high as 2,031 CFU/ml. Counts for individual towers also varied greatly, with standard deviations generally three times as large as the mean. Time series correlograms were used to determine whether significant correlation existed between each sample and samples taken subsequently from the same tower.

Significant correlation between one sample and samples taken less than one week later, was observed for only 9 towers out of 28. After a 2 week interval, only 3 towers showed significant correlation between samples, and after 3 weeks, only one tower showed significant correlation. Therefore, given the time delay of 7-10 days to obtain a culture result, the number of Legionella present in a cooling tower will probably have changed significantly by the time the results of a sample are obtained.

Examination of the range of cell counts for individual towers showed that only 1 would have been classified as high risk on the basis of mean Legionella count (mean >1,000 CFU/ml), however 10 systems had individual counts in this range at least once during the 16-week sampling period. All systems showed Legionella counts in the low risk range (<100 CFU/ml) at some time during the study.

The results of this study demonstrate that single samples or infrequent sampling for Legionella in cooling towers is unlikely to provide a meaningful measure of the potential infection risk. The relationship between Legionella numbers and infection risks is also complicated by variations in the infectivity and virulence of different species and strains of the organism.

The practice of assigning the probable infection source on the basis of DNA fingerprinting of isolates has also been thrown into doubt by other recently published work on Legionella typing⁽³⁾. Outbreak investigation generally involves the comparison of human clinical isolates with those taken from environmental sources, using a variety of serological, biochemical or molecular genetic tests. Identical results in such tests are often interpreted as evidence that two or more isolates are the same, and that a particular environmental source is the likely origin of the outbreak.

Pulsed Field Gel Electrophoresis (PFGE) is one of the methods used in this kind of investigation. This molecular genetic technique involves the digestion of genetic material with restriction enzymes and separation of the resultant fragments on agarose gels. Researchers from the US and Canada compared PFGE patterns from 62 L. pneumophila isolates collected over a 14 year period from 9 states, and found that common PFGE patterns occurred among a number of geographically and temporaly distinct isolates. This suggests that some strains are widely dispersed and persistent in the environment, and that the simplistic assumption that a particular cooling tower or other location is the source of an outbreak on the basis of limited characterisation is not valid. A combination of analytical techniques together with knowledge of the frequency and distribution of different strains is required to allow more certainty in assigning the source of outbreaks.

The Walkerton Inquiry has completed the first phase of its investigation (Part Ia) into the waterborne outbreak of E. coli and Campylobacter infections that hit the small town of Walkerton, Ontario in May 2000. In the period from October 2000 to January 2001 the Inquiry heard testimony from 44 people and received over 1,300 documents in evidence regarding the circumstances of the outbreak in which 7 people died and over 2,300 became ill. Witnesses included the former Manager of the Public Utilities Commission, who admitted that he had frequently falsified chlorination readings, and that he and his staff had little training in drinking water quality management.

In a statement issued at the close of the first phase of the Inquiry, Commissioner Dennis R O'Connor said that he was pleased with the progress of the Inquiry, and satisfied that he had the necessary evidence to make the required findings in respect of the circumstances which caused the outbreak. The Commissioner also praised the conduct of the legal counsel for the various parties in avoiding procedural delays and arguments, and enabling the Inquiry to remain on schedule.

The second phase of the Inquiry (Part Ib) will examine the effect, if any, of government policies, practices and procedures on the events in Walkerton. This will commence with the review of over 100,000 documents from a wide range of government ministries and agencies, followed by more public hearings. The Inquiry will also receive advice from its Scientific Advisory Panel on the physical cause of the contamination in Walkerton.

Meanwhile, Part II of the Inquiry concerning the future management of Ontario's water supplies has already commenced, with the commissioning of a number of expert reports. The Inquiry will also hold public meetings in several towns across Ontario to facilitate input from a wide range of people and organisations. Part II is scheduled for completion in September 2001, with the final Report of the Inquiry to be delivered to the Ontario Attorney General by the end of the year.

Negotiations between the Ontario government and representatives of Walkerton residents have resulted in a "no-fault" settlement of the legal class action brought as a result of the outbreak. Under the terms of the Agreement, ratified by the Chief Justice of the Ontario Superior Court on 19 March, residents and visitors who suffered illness as a result of the water contamination will receive a minimum payout of $2,000. Additional claims may also be made for the illness or death of a relative, and for economic losses relating to the outbreak. Part of the costs for the compensation payments will be shared by the insurers of Walkerton's Public Utilities Commission, who will pay legal fees of $4 million for the plaintiff’s lawyers and the first $17 million of claims. The reminder of the compensation costs will be met by the Ontario provincial government.

Arrangements for meeting the $12 million costs of decontaminating the Walkerton water supply and leasing a temporary filtration system have not been finalised. A preliminary report from government appointed auditors has suggested that the Walkerton community meet $3 million of the cost plus a $1.5 million capital contribution to establishing a long term alternative water supply, however negotiations between the local council and the Department of Municipal affairs are continuing. The council has also raised concerns that a planned health study of Walkerton residents has still not been funded some five months after it was announced by the Ontario Minister for Health.

US researchers recently reported they had determined the sequence of most of the E. coli O157:H7 genome. This enterotoxin producing E. coli and other related strains have been the cause of a number of waterborne disease outbreaks (including the Walkerton outbreak in Canada last year). It is already known that toxin producing strains carry additional genes specifying their characteristic toxins and several virulence factors which are not possessed by the more common non-pathogenic E. coli strains, however the researchers were surprised to find the extent of genetic differences revealed by the genomic sequence comparison.

The E. coli O157:H7 type strain (originating from a 1982 ground beef food poisoning outbreak) contains an estimated 1,387 genes that are not carried by the non-pathogenic laboratory strain E. coli K-12, making its genome about 32% larger. The additional genetic material is interspersed in at least 177 positions around the single circular chromosome, suggesting that genetic material has been acquired or lost in many separate events since the two strains diverged an estimated 4.5 million years ago.

The additional material is far in excess of that required to encode the known virulence determinants of the pathogenic strain, and may contain genes for as yet unidentified virulence characteristicss, additional metabolic capabilities or latent bacteriophages.

The UK Department of Environment, Transport and Regions has announced the result of an 18-month study of radon levels in 116 private water supplies in the west Devon region. About 18% of homes in this area use private wells or bores for drinking water. The study showed that about 1 in 7 homes had radon levels above the draft European Union Commission Recommendation action level of 1000Bq/l, and 1 in 14 had uranium levels above the WHO provisional guideline of 2 microgram/L.

The study also included laboratory studies of radon loss during the preparation of hot and cold drinks, and found that losses were less than previously reported in the literature. Estimates of individual ingested doses of radon showed that some people might receive an exposure of over 13 mSv per annum. Exposure from inhalation may be higher than that from ingestion. Current regulations governing radiation workers require safety monitoring, protective equipment and medical checks for those receiving doses of 6 mSv or more.

The local council has called for a larger study in all radon affected areas of Britain, while public health authorities have advised consumers not to be unnecessarily concerned as guideline levels include wide safety margins.

The Prime Minister of China recently announced that plans to build a major pipeline to transport water from the rivers and lakes in southern China to the arid north would be accelerated. It is estimated that two-thirds of China’s 600 cities suffer from water shortages, and limitations in supply are seen as a significant obstacle to further economic development. The pipeline project involves three networks of aqueducts, canals and pipes with a combined length of over 2,1000 miles which are designed to carry 1.87 trillion cubic feet of water per year.

While the latest 5-year plan for economic development calls for more emphasis on sustainable development strategies, environmental advocates still fear that inadequate government controls and entrenched corruption will see continuation of widespread ecological destruction. They warn that the pipeline plan may result in southern rivers running dry.

Jail for Selling Tap Water
A British business man was recently jailed for 6 months for selling bottled tap water as spring water. The former winner of a Businessman of the Year Award pleaded guilty to fraudulent trading and acting as a company director while being an undischarged bankrupt. His company, which operated for only 9 months, initially bottled genuine spring water from rural Cumbria, but when the supply became inadequate he ordered employees to use water from a number of different sources including the ordinary tap water supply.

Contact Information

The printed version of Health Stream is available free of charge - to be added to our mailing list please contact Pam Lightbody (email or fax + 61 3 9903 0576). Past issues can be found under Publications, and there is a searchable archive of articles, news items and literature summaries.